Questions

How to model low (planetary boundary layer) clouds?
How to think about the probability distribution of cloud types, and how might it shift with climate change?
How to model the fraction of cloud cover and what are better parameterizations? Total cloud albedo is amazingly constant -- why?
How to combine models with a good vertical resolution with an account of micro and macro-turbulence?
Can cloud cover be viewed as a problem of optimization? Maximum entropy ideas?
To what degree are precipitation extremes controlled by properties of clouds?
Aerosols and cloud nucleation processes?

Are in-situ experiments that modify temperature, CO2, and/or moisture realistic simulations of a changing climate?
How can carbon-cycle models be verified? Over what time and spatial scales?
What is the best model of the biosphere that we can realistically hope for?
Measurements of CO2 in a vertical column are needed (large diurnal oscillations in surface CO2 concentrations due to plant photosynthesis / respiration).
Do we need to worry about species instead of just plant functional types?

Pressing need to model ice flow over surfaces (ie. the Greenland and Antarctic ice sheets). How can this be done?

Do statistical closures exist that reproduce, at least qualitatively, the main features of extratropical circulation? Quantitatively?
What is the basic mechanism of Madden-Julian oscillations that determines their period and the speed of propagation?
Origin of power law scaling in mesoscale turbulence.
Moisture: Effect of water vapor and latent heat on large-scale eddies.
Simple models of the monsoon -- can they be constructed? What are the basic drivers?

Are there locations where the formation of deep water could keep CO2 away from the atmosphere for > 100 years?
To what degree is mixing a function of climate? What sets rate of vertical mixing rates? What controls heat transport?